Tobacco smoke is the single most preventable risk of cancer and is responsible for lung cancer being the highest in incidence worldwide. Research on the association between tobacco and cancer relied on questionnaire data for tobacco exposure assessment which are subjective and inaccurate. Therefore, the true risk of cancer from tobacco exposure is expected to be higher than what is reported in the literature and would lead to null findings for low risks. Biomarkers of tobacco exposure offer complementary objective measures that can improve exposure assessment of tobacco. Some of the obstacles to wider use of biomarkers are the lack of population based reference values for determination of standards of exposure or cut-off points in relation to health outcomes. For tobacco exposure, serum, saliva and urine cotinine biomarkers have been more widely used than others because population values do exist for them. However, these biomarkers are influsure. However, little is known about the population levels of hair and toenail nicotine levels and the overall variability among healthy individuals in the general population. The aim in this application is to assess mean population levels of hair and toenail nicotine and their variability in a healthy general population in California participating in the California Tobacco Survey. A total of 500 participants will be randomly selected from the 2008 Survey and interviewed by phone on tobacco exposure related questions and hair and toenail samples collected by mail. A subsample of 100 participants will be asked to send salivary cotinine samples to be compared to toenail and hair nicotine levels for sensitivity and specificity purposes. We will also assess questionnaire variables that best predict these biomarkers to determine suitability for cancer related studies. Results from this study would help standardize these biomarker levels and provide reference values for wider future use and s will assist in refining the prevalence of smoking for the California Tobacco Survey and other population surveys as well as improve the prediction of cancer risk.